Evaluating the Impact of Entrance Awards on Yield Rates Using Machine Learning

Entrance awards are financial incentives offered to admitted students at or near the time of admission. These awards are a frequently used tool to enhance yield rates among post-secondary institutions, sometimes among specifically-targeted populations. The impacts of these awards have been studied in many different contexts, using a variety of methods (Bartik, Hershbein and Lachowska 2021, Cornwell, Mustard and Sridhar 2006, Crowne 2022, Gurantz and Odle 2022). The most effective way to quantify the effect of entrance awards is to assign them randomly among qualified applicants, and then compare the outcomes with the outcome of the group that was not assigned any awards. This method was used in (Firoozi 2022) but is not advisable in the authors’ regulatory environment. Therefore, this research compared the observed yield rate among students to the yield rate that was predicted using a machine learning model that was trained on data collected before the introduction of entrance awards. The predicted yield rates can be thought of as the behaviour of a “synthetic control group” that behaves as would be expected if no entrance awards were offered, allowing the researchers to measure the impact of the awards.